万卡集群，郑州抢跑！国家超算互联网核心节点上线试运行

Core Insights - The National Supercomputing Internet Core Node has commenced trial operations, supported by the Shuguang scaleX WanKa supercluster system, providing over 30,000 domestic AI computing resources, marking the largest single domestic AI computing resource pool connected to the National Supercomputing Internet platform [1][3] - The core node aims to unify computing power distribution, effectively connecting supply and demand, and promoting the construction of an AI application ecosystem [3] - Zhengzhou's core node is designed for flexible scalability, capable of expanding to 100,000 or even 1,000,000 computing units, and is strategically located to enhance regional development [3][4] Infrastructure and Development - The core node will significantly enhance high-performance computing capabilities in Zhengzhou and the central region, facilitating nationwide unified scheduling of computing resources through high-speed communication networks [4] - Zhengzhou plans to integrate supercomputing resources with local industries, focusing on sectors such as intelligent sensors, new energy vehicles, high-end equipment, and new materials [4] - The establishment of a computing power trading system and emphasis on data security and privacy protection are also part of Zhengzhou's strategy to expand local computing service applications [4] Application and User Engagement - The supercomputing internet is transitioning to a model that emphasizes both construction and application, aiming to create a marketplace for computing applications similar to e-commerce platforms [5] - By the end of 2025, the platform is expected to serve over 1 million users, with more than 7,200 application products and a peak daily processing workload of 1.03 million tasks [5] - The core node's testing program includes lightweight trials and large-scale testing, offering resources for various user groups, including startups and students [7] Innovation and Research - The core node has achieved innovative applications in key areas such as artificial intelligence, watershed management, weather forecasting, molecular simulation, and industrial simulation, with over 400 mainstream models optimized [7] - The platform aims to provide efficient and accessible computing services for various cutting-edge application scenarios, promoting the free flow and usability of computing power [7]